Poly-GAN: Regularizing Polygons with Generative Adversarial Networks

Regularizing polygons involves simplifying irregular and noisy shapes of built environment objects (e.g. buildings) to ensure that they are accurately represented using a minimum number of vertices. It is a vital processing step when creating/transmitting online digital maps so that they occupy minimal storage space and bandwidth. This paper presents a data-driven and Deep Learning (DL) based approach for regularizing OpenStreetMap building polygon edges. The study introduces a building footprint regularization technique (Poly-GAN) that utilises a Generative Adversarial Network model trained on irregular building footprints and OSM vector data. The proposed method is particularly relevant for map features predicted by Machine Learning (ML) algorithms in the GIScience domain, where information overload remains a significant problem in many cartographic/LBS applications. It addresses the limitations of traditional cartographic regularization/generalization algorithms, which can struggle with producing both accurate and minimal representations of multisided built environment objects. Furthermore, future work proposes a way to test the method on even more complex object shapes to address this limitation.